As other reviewers have mentioned, the configuration possibilities are endless in terms of how you can set up your subscriptions, the payment gateways you can use, and the integrations to other components really give you some power. The sheer fact that I can fire off just about any kind of action (db, email, web, sms) from a custom script in response to subscription events makes this a very powerful tool.
But AEC's greatest strength is also it's weakness - vast configuration. So much time has been spent trying to cover every angle that alot of polish was missed on just getting the very basics solid.
Here's an example, in the US, Authorize.net is utilized heavily. AEC recently moved from using Auth.net ARB based subscriptions to the more powerful CIM system at Auth.net - a good move in my view, but the implementation was very sloppy. The "happy path" works for checking out, but things get confusing when you run scenarios like a "failed" credit card at checkout, or incorrect handling of the messaging coming back from auth.net - like a cv2 code problem, or mismatched billing address (if you're using that verification in Auth.net, which you certainly should be). AEC also doesn't utilize the "preauth" feature (or at least offer it as an option) with CIM where you can validate the card at checkout time - you could as a result end up giving a customer access to a paid service until the first payment is attempted on the billing cycle. Offer a trial, and it compounds the issue. On failed transactions at checkout, the screen that comes back is insanely confusing to the end user. It displays the "profile" that was created in Auth.net (why do they need to see this? The customer doesn't care) with a cryptic message about having to re-enter their cv2 code and credit card number if they are changing their profile..."huh? profile? I just want to check out man!" This is where alot of the polish is missing - an attractive, but solid and functional end-user interface that handles all the permutations. The basics. But - I can sure create "fallback" subscription rules and "subscription limits" and coupons to my heart's content....if anyone's using that. My opinion only of course, but those advanced features are nice to have and should be developed after the basics are good and solid. I suppose that having to support so many checkout gateways, causes this problem as it's pretty common in most of the subscription management components out there - lots of the "basics" of operation just miss.
Aside from that the other real problem here is the business model. As others have stated this just has to be thrown out if the masses are going to take AEC seriously. I have no problem paying, and continuing to pay for upgrades and direct support like any other software model. AEC is mature enough that it's worth paying for. It's absolutely maddening though that I have to keep paying to get access to the documentation for a version I've already paid for, and especially access to a forum that is end-user driven ONLY. I lost count of the times I ran into an issue (i.e., custom code cleanup to fix that end-user interface html 101 stuff - come on guys, get a designer!) and wanted to visit the forum to see if any other users had suggestions or similar problems, only to find that my subscription had "just" run out. This is probably the single reason we finally walked from AEC and went with another tool. That's just absolutely re-donk-ulous.
As an aside, the ultimate future and direction of AEC is/was really unclear. David, AEC's creator (I believe) often saw it fit to turn the AEC site into a soap box for complaining about technology this, and process that, and often wrote long essays on how depressed he was and needed a vacation and yadda yadda. The problem is, you're users see this and wonder "ok, is AEC dying? Is this project stable? We've all invested alot of time into this so are you going to stop working on some of these issues that are so close to really working well". Really unprofessional in my view. Take the rants to Facebook - and keep the AEC site focused on AEC.
Aside from those negative points, again AEC is probably the most advanced and moldable Joomla paid subscription manager out there. Some of the bits are in fact REALLY well thought out, and you can see that when you dig into the code. I won't say it's the best solution out there - because everyone's needs are different. There are simpler and more attractive components out there if you're needs are simple, and you need it to "just work".
If you're going to use AEC, prepare for a learning curve, and have some html/php/Joomla skills on hand. It's really got the potential to be THE BEST component in this class, if Valanx and crew get that business model more realistic (even if it ultimately costs the end user MORE - let me pay for software upgrades and support - not documentation and end-user forums!), get some of the permutations with checkout and profile management in line, and get an artistic designer in there to fix the front end blah. It's close guys! Help me come back to you -